Validation of the Anxiety Intensity Scale Circles (AISCs) and Yale-A

  • Research type

    Research Study

  • Full title

    Anxiety Intensity Scale Circles (AISCs) and the Yale-anxiety: exploring the validity of two anxiety screens and Natural Language Processing in stroke

  • IRAS ID

    326826

  • Contact name

    Joshua Blake

  • Contact email

    joshua.blake@uea.ac.uk

  • Sponsor organisation

    University of East Anglia

  • Duration of Study in the UK

    1 years, 3 months, 1 days

  • Research summary

    Feeling anxious is common after stroke. It is important that clinicians are aware if someone is suffering from anxiety so that they can provide the best support.

    Clinicians use questionnaires to understand if people are feeling anxious. However, problems with language, attention, or fatigue, after stroke can make it hard for some stroke survivors to complete. We therefore aim to develop new questionnaires that people with these problems can complete, so that their experiences are not overlooked. The questionnaires are the Anxiety Intensity Scale Circles (AISCs) and the Yale question for anxiety (Yale-a). They are designed to be more inclusive for people with the above problems than other measures by relying less on language, being briefer, and easier to answer.

    We aim to identify if these two measures are accurate. We will ask 142 stroke survivors on rehabilitation wards, at least two weeks after their stroke to take part. They will be asked to complete the AISCs and Yale-a alongside some extra mood and anxiety questionnaires. We will assess if the scores are similar (convergent validity) and compare completion rates. A mental health professional will then talk to the participants separately and conduct a clinical interview. If the results of the AISCs and Yale-a match the result of the clinical interview, we will know they are accurate (criterion validity). Finally, we will repeat the questionnaires on a third visit to see if the results stay similar over time.

    We also want to see if Artificial Intelligence (AI) can detect features of anxiety and low mood in people's tone of voice and the words they choose. We will audio record two description tasks and train the AI using the clinical interview outcome data.

  • REC name

    London - Hampstead Research Ethics Committee

  • REC reference

    25/LO/0456

  • Date of REC Opinion

    11 Jul 2025

  • REC opinion

    Further Information Favourable Opinion